"Moderates, those with a prior of 4, made the most valid model choice and are therefore most likely to obtain the correct estimate. Hence, pro-immigration researchers in particular appear to have reached their conclusion by making less valid choices in their analysis, thereby reaching a biased conclusion. If anti- and pro-immigration researchers were similarly frequent, such a bias may cancel out. However, pro-immigration researchers outnumber anti-immigration researchers and moderates, whose analyses are most reliable combined. The results is a literature containing mainly a small share of moderates doing rigorous research and a larger group of ideologically motivated pro-immigration “researchers” who torture the data until they “find” a pro-immigration message."
laurenzguenther.substack.com/p…
Why immigration research is probably biased
Public opinion surveys across Western countries consistently show that majorities prefer lower immigration levels.Laurenz Guenther
miki
in reply to J. B. Crawford • • •A thought I've had previously:
How long would it take a group of motivated individuals to make a modern computer, if they were sent to Earth as it existed in e.g. the 18th century, and given an artifact to communicate with Earth as it exists now.